Credits

Contact

Deleting your account is permanent. Contact support to close your account.

Delete account

Crop image

Quick links

Require a one-time code from an authenticator app when you sign in.

Set up two-factor authentication

Two-factor authentication

Not signed in

Overview

Name

User ID

Profile

Manage your password, email, and two-factor authentication.

You signed in with Google or Apple, so your email, password, and two-factor settings are managed there.

Managed by your sign-in provider

Security

Enter the 6-digit code we sent to your new address.

Verify your new email

Your profile and credit summary.

Account

Signing you in…

If an account exists for {email}, a reset code is on its way.

Check your email

We'll email you a reset code.

Reset your password

Enter the 6-digit code from your authenticator app.

Your account needs a new password before you can continue.

Set a new password

Enter the code from your email and choose a new password.

Access your research workspace.

Sign in

Start your research workspace.

Create your account

Enter the code we sent to {email}.

Verify your email

Trading strategy research, backtesting, and optimization.

StatePath Research

This conversation reached its length limit. Start a new session to continue.

This content has been moderated and can no longer be edited.

This content is licensed sealed: you can run it, but its internals stay with the seller.

Something broke

Couldn't load this resource

Run failed

Now you try

Key terms

Learn to design trading strategies on the visual canvas — step by step, hands-on, in the real editor.

Learn · StatePath

Learn to design a strategy

Pass only when ADX(14) is above 25 — keeps trend strategies out of chop.

Trend Strength Gate

Size the stop and target to volatility — one ATR(14) either side of price.

ATR Volatility Bracket

Trade momentum shifts as the Awesome Oscillator crosses its zero line.

Awesome Zero Cross

Ride expansion: buy when price breaks the upper band, sell when it breaks the lower.

Bollinger Breakout

A classic ±100 mean-reversion trigger on the Commodity Channel Index.

CCI Reversal

The turtle channel breakout — buy the 20-bar high, sell the 20-bar low.

Donchian Breakout

Buy when the fast EMA crosses above the slow EMA; sell on the cross down. The classic trend-following entry.

EMA Crossover

Enter on the Tenkan/Kijun cross — Ichimoku's core trend trigger.

Ichimoku TK Cross

Confirm longs only when the MACD histogram is positive, shorts only when negative.

MACD Momentum Gate

Enter when the MACD line crosses its signal line — momentum turning with the trend.

MACD Signal Cross

A quality gate that blocks entries while RSI(14) is above 70 — don't chase overbought markets.

Not Overbought

Fade extremes: buy when RSI(14) is oversold below 30, sell when overbought above 70.

RSI Reversal

Only trade the liquid 08:00–16:00 window; skip the thin overnight session.

Session Hours

The smoothed-momentum cousin of RSI reversal — buy below 20, sell above 80.

Stochastic RSI Reversal

Place the stop under the most recent swing low from a zigzag pivot, with a fixed take-profit.

Swing-Low Stop

Recipe Library

Ready-made Search Signals, Filters and Create Orders to copy into the designer and adapt — across trend, momentum, mean-reversion, breakout and more.

Recipe Library · StatePath

With a signal, a gate and an order published, the Composer snaps them onto the pipeline board — Search Signal on the left, States in the middle, Create Order on the right — no wiring. The result is one strategy you can validate as a whole.

Stage 4 — assemble & publish

A raw signal fires too often. Wrap it in States of filters to keep only the good ones: a trend-strength gate to skip chop, a session-hours gate to skip thin liquidity, a not-overbought gate to avoid chasing. Each is a small Filter document dropped into a Strict or Non-Strict zone.

Stage 2 — refine with filters

You’ve built each piece in isolation. A real strategy is those pieces wired into one pipeline — a fixed assembly line every signal travels down.

From tools to a tradeable system

That’s the whole loop: spark, judge, execute, manage. Every strategy in the Recipe Library is just these four stages with different tools inside. Build one, publish it, backtest it — then come back and tune.

Where to go next

Stage 3 — set the risk

Start with what sparks a trade. A Search Signal watches the market and fires a raw <code>buy</code> or <code>sell</code> — nothing about risk yet, just something worth looking at happened. The trend-following classic is a moving-average crossover.

Stage 1 — the entry trigger

Snap the pieces into the four-stage pipeline — signal, filters, order — and ship a tradeable system.

Composing a full strategy

In Foundations you met a State as a checkpoint returning <go>GO_NEXT</go>, <hold>HOLD</hold> or <cancel>CANCEL</cancel>. Now look inside one. A State holds one ordered lane of filter slots, walked top to bottom — no buckets, no groups, no score to tune.

A State is a lane, not a bucket

Open a State on the composer board and you see a numbered lane: slot 1, slot 2, slot 3. The engine walks them in exactly that order. Drag the <code>⠿</code> grip to reorder a slot and the numbers — and the evaluation order — follow. The same filter may appear twice in one lane; a slot is addressed by its position, not by which filter sits in it.

One ordered lane

Each slot reports <pass>PASS</pass>, <hold>HOLD</hold> or <cancel>CANCEL</cancel> for a direction. The State folds the outcomes of its gates into a single verdict:

A gate wired its <code>cancel_buy</code> / <code>cancel_sell</code> exit and it fired. The signal is discarded for good — it will not be reconsidered on later bars.

Any gate cancelled

All gates approved this direction. The signal advances to the next State, or out of the stage entirely.

Every gate passed

No gate cancelled, but at least one did not pass. The signal is parked and re-evaluated as new bars arrive, until every gate passes or the hold timeout expires.

A gate is still waiting

Three outcomes, one verdict

One ordered lane. Three outcomes. Gates vote, observers only watch. A cancel stops everything below it; a miss simply waits. Here are two ready-made filter recipes to drop into a lane.

Journey, recap & recipes

Every slot carries a role. A gate is the default: its outcome is folded into the verdict. Deal-breakers and confirmations are both gates — what separates them is how the filter wires its own exits, not which bucket it sits in.

Gates and observers

Worked example: a band-touch filter

One ordered lane of filters. Three outcomes. And who actually gets a vote.

Assembling a State

Assembling States from ordered filter lanes, and composing a full strategy.

Advanced

Every input gets its value one of three ways — picked with a switch in the Inspector. A port only shows a wire-handle in <code>Canvas</code> mode.

Three ways to feed an input

No code, no jargon. A strategy document is a small diagram you build by hand — boxes that each do one job, joined by wires. Three moves are all the mechanics you need.

How a document is built

You’ve met the canvas: tools become nodes, wires carry typed data left → right, inputs are Fixed / Module / Canvas, and every document ends at one Exit. That is the same surface for all three editors.

Recap

A port’s shape tells you the kind of data; its colour tells you the flavour. Matching shapes connect cleanly.

The shape-and-colour language of wires

Worked example: the moving-average crossover

Every editor works the same way — learn the canvas once and you can build all three.

The canvas, tools & wires

Click <code>+ Add state</code> for multi-stage logic — give it a name like <code>confirm</code>. States evaluate left to right, exactly the chaining from the filters lesson. The first state is always <code>Default</code> — it can’t be renamed or removed.

Add states, run left → right

You’ve now seen every piece: a spark that finds opportunities, States of Strict & Non-Strict filters that judge them, and a Create Order that turns the survivors into trades — assembled on one board.

Search Signal → States → Create Order

Drop, don’t wire

Three building blocks become one strategy. The Composer is a fixed pipeline board — you drop published documents into slots rather than drawing wires.

Snapping the pieces together

Worked example: assemble the pipeline

Three building blocks become one strategy — drop published documents into a fixed pipeline.

The Composer

A signal stage isn’t limited to one checkpoint. Chain several States and a signal must clear them in order. <go>GO_NEXT</go> advances immediately — it can clear several green checkpoints on the same bar. <hold>HOLD</hold> parks the signal at that State; it’s re-checked there on later bars. <cancel>CANCEL</cancel> discards it on the spot. The very first State is always named <code>Default</code>.

Multiple States in a row

Watch real signals flow through a two-State pipeline. Click any signal in the table to see each slot's outcome and which gate decided its fate. Both lanes hold gates — they differ only in whether those gates wired their cancel ports.

Trace a signal through the States

It never creates a signal — it judges one. Its Exit box, <code>FILTER_EXIT</code>, has four yes/no (◆) ports: <code>buy</code> and <code>sell</code> approve a direction, <code>cancel_buy</code> and <code>cancel_sell</code> veto one.

A filter is a quality gate

A Search Signal spots opportunities; filters decide which ones are worth trading. This is where most of your skill as a strategy designer lives.

Where your design skill lives

One filter is rarely enough. Filters live inside States — checkpoints holding one ordered lane, which return one of three verdicts. Each slot carries a role that decides whether its outcome counts.

Combining filters: States, gates & observers

A gate vetoed this direction — it wired its <code>cancel_buy</code> / <code>cancel_sell</code> exit and it fired. The signal is thrown away for good — it will not be reconsidered on later bars.

Discard

Every gate passed — the signal moves to the next checkpoint, or out of the stage. On the same bar it can clear several green checkpoints at once.

Advance

No gate cancelled, but at least one did not pass. The signal is parked and re-checked on later bars — not killed. The moment every gate passes, it advances.

Wait & retry

Worked example: only trade when RSI is below 70

A Search Signal spots opportunities; filters decide which ones are worth trading.

Filters, outcomes & States

A signal that cleared every checkpoint is approved — now it becomes a real order. The Create Order’s only job: where do the stop and target go?

Turning a signal into an order

Look closely at ORDER EXIT: every slot exposes a buy and a sell port — <code>sl_1_buy</code> / <code>sl_1_sell</code>, <code>tp_1_buy</code> / <code>tp_1_sell</code>. That lets you set a different stop or target for longs vs shorts — a swing-low stop for buys, a swing-high stop for sells. For a long-only order you wire only the <code>_buy</code> side — which is exactly what the worked example does.

Each slot has a buy and a sell side

TP and SL are actual prices (like 1.1050), not distances. The slots want a single number (●), so you feed them from <code>constant.number</code> (a fixed price you type in) or <code>interpretor.point</code> (a swing high/low price off the chart).

Feed it prices (●)

What it sets — and what it inherits

Worked example: stop under the swing low

A signal that cleared every checkpoint is approved — now it becomes a real order.

The Create Order

The canvas, signals, filters and orders — start here.

Foundations

Ready-made strategies, filters and signals to copy and adapt.

We're doing scheduled maintenance. Everything still works — you may see brief interruptions.

Scheduled maintenance

The trading platform is temporarily unavailable while we work on it. Your data is safe and nothing is lost.

We'll be back shortly

Command palette

Short, interactive walkthroughs of each part of the platform. Start one anytime — your progress is saved.